Invasive Filamentous Fungal Disease: ICU Diagnosis, Therapy
Invasive filamentous fungal disease in critically ill patients has become one of the hardest infectious problems in the intensive care unit. Aspergillus, the Mucorales order and Fusarium account for most cases, and the patients who develop them are no longer only those with low neutrophil counts. The disease is fast, the signs are vague, and the mortality runs from 40 percent to over 80 percent depending on the pathogen. This article explains what the disease is, who is at risk in the ICU, how the diagnosis is built, and how the main antifungal regimens are chosen and adjusted.
Quick answer
Invasive filamentous fungal disease in critically ill patients is driven mainly by Aspergillus, Mucorales and Fusarium. Mortality runs from 40 to 82 percent by pathogen. Diagnosis combines host risk, radiology, direct microscopy, culture, PCR and galactomannan from bronchoalveolar lavage. First line therapy is isavuconazole or voriconazole for aspergillosis and liposomal amphotericin B for mucormycosis, with early surgery where a lesion can be removed.
Why this matters in the ICU
Critically ill patients in the ICU carry a heavy burden of filamentous fungal infection, and the problem keeps growing. Several forces push it up at once: newer and more aggressive immunosuppressive and targeted drugs, broader use of broad spectrum antibiotics, more organ transplants, and the waves of viral pneumonias that have swept through hospitals. The lung is the organ most often involved, in more than 60 percent of cases.
The patients who get sick are a mixed group. Many are not neutropenic. They have severe underlying disease, they are on mechanical support, and they receive stacked therapies that each nudge the immune system down a little further. That makes the picture different from the classic neutropenic haematology patient, and it changes both how you suspect the disease and how you treat it.
The practical danger is simple. The signs, fever, cough, falling oxygen, new infiltrate, are exactly the signs every ICU patient already has for other reasons. So the disease hides inside the routine. Early recognition is the single factor that moves survival.
The pathogens and how often they kill
Filamentous fungi are environmental moulds that most healthy people inhale without trouble. In a weakened host they invade tissue and blood vessels. Three groups dominate the ICU.
| Pathogen group | Common species | Share of disease | Reported mortality in critically ill patients |
| Aspergillus | Aspergillus fumigatus 50 to 90 percent, A. flavus 15 to 25 percent, A. niger 5 to 10 percent, A. terreus under 5 percent | Most common cause of invasive filamentous fungal disease | 46 to 82 percent for invasive aspergillosis |
| Mucorales | Rhizopus, Lichtheimia, Rhizomucor | Causes rhino orbital cerebral, pulmonary and cutaneous disease | Up to 50 percent; global prevalence 0.005 to 1.7 per million |
| Fusarium | Fusarium solani complex about 50 percent, F. oxysporum complex about 20 percent | Often disseminated with fungaemia | 40 to 70 percent |
Some species carry built in resistance. Aspergillus terreus is naturally resistant to amphotericin B. The Mucorales order is naturally resistant to voriconazole. Parts of Fusarium and Scedosporium are also poorly served by amphotericin B. Knowing the likely species from the host and the site tells you which drug to avoid before the culture comes back.
Acquired azole resistance in Aspergillus is a real and growing problem. It travels through CYP51A mutations, efflux pumps and stress pathways. Reported azole resistant Aspergillus stays under 10 percent of isolates in most of the world, with regional figures reported as 1.4 to 5.6 percent in China and a rising trend.
Risk factors in non neutropenic ICU patients
The risk map for non neutropenic critically ill patients is its own thing. It is built from the host disease, the treatments given in the ICU, and the environment.
Influenza and COVID 19 deserve special attention because they drive influenza associated pulmonary aspergillosis (IAPA) and COVID 19 associated pulmonary aspergillosis (CAPA), two forms that strike otherwise non neutropenic ventilated patients.
| Fungus | Host factors | Medical interventions | Environmental exposure |
| Aspergillus | Moderate to severe COPD, severe influenza, severe COVID 19, uncontrolled HIV with CD4 under 200 per cubic millimetre, solid tumours, decompensated cirrhosis, solid organ transplant | Long term high dose glucocorticoids, for example prednisone 20 mg per day or more for over 2 weeks or over 2 g cumulative in 3 months, calcineurin or mTOR inhibitors within the prior 90 days | High concentration mould spore environments such as construction work or potted soil |
| Mucorales | Poorly controlled diabetes, especially with ketoacidosis or hyperosmolar crisis, severe trauma or burns, severe COVID 19, solid organ transplant, iron overload | Soil contaminated wounds from trauma, contaminated air handling systems | |
| Fusarium | Severe COVID 19, severe trauma or burns, solid organ transplant | Drowning or injury in swamp or sewage contaminated environments |
Two points deserve emphasis. First, glucocorticoids at the doses used in severe COVID 19 and severe asthma are a central driver of Aspergillus in patients who are not neutropenic. Second, uncontrolled diabetes with ketoacidosis is the main doorway to Mucorales. One third of adult Fusarium patients are immunocompetent, which means a normal immune system is no guarantee against that pathogen.
How the disease presents
Filamentous fungal disease usually starts fast and moves fast. It tends to cause tissue necrosis, blood vessel invasion and spread. The picture tracks the pathogen and the site, but none of it is specific.
Invasive aspergillosis most often hits the lung. Non neutropenic patients frequently show an airway invasive pattern early: persistent cough, wheeze, mucus plugs and breathlessness, with invasive tracheobronchial aspergillosis a real possibility. Sinus and orbital involvement brings facial pain, swelling, nosebleed and loss of smell. Central nervous system involvement brings headache, seizures and focal deficits. Disseminated disease produces multiorgan failure.
Mucormycosis is defined by its vascular aggression. Rhino orbital cerebral disease starts with fever, face pain and black nasal discharge, then moves to the eye and brain. Pulmonary mucormycosis causes severe haemoptysis because it destroys vessels. Gastrointestinal and disseminated forms follow contaminated wounds or uncontrolled diabetes.
Invasive fusariosis looks like pulmonary aspergillosis in the lung, but it stands out for fungaemia, which occurs in 40 to 60 percent of cases, a rate well above the other filamentous fungi, and it spreads to bone, joint and brain.
Bronchoscopy, imaging and tissue
Bronchoscopy earns its place in the ICU because it lets you see the airway and take a targeted sample. Invasive tracheobronchial aspergillosis shows as ulcers, pseudomembranes or black eschars. Ulcers are clean bordered mucosal defects. Pseudomembranes are grey or yellow white coatings that can block the lumen. Eschars, the black dry necrotic plaques, are most suggestive of Mucorales. Biopsy of an ulcer can confirm invasive disease. Bronchoalveolar lavage fluid is the sample that powers the best non culture tests.
Imaging overlaps between the three fungi, so it rarely names the species alone. Useful signs:
| Pathogen | Site | Imaging signal |
| Aspergillus | Airways | Tree in bud and centrilobular nodules mark airway invasion |
| Aspergillus | Lung | Patchy consolidation, nodules along vessels, necrosis; halo and air crescent are uncommon, under 20 percent |
| Mucorales | Lung | Thick wall irregular cavity; the reversed halo sign early is relatively characteristic |
| Fusarium | Lung | Mostly nodules, halo rare, tree in bud usually absent |
Tissue pathology remains the proof of invasion. Aspergillus hyphae are thin, about 3 to 5 micrometre, septate and branch at acute angles. Mucorales hyphae are broad, about 5 to 25 micrometre, sparsely septate and branch at right angles, with worse vascular invasion, thrombosis and haemorrhagic necrosis. Pathology confirms invasion and points to the family but usually cannot name the exact species, and it cannot separate Aspergillus from Fusarium or Scedosporium.
Getting tissue is hard in the critically ill. Take it early when it is safe, send it with the microbiology sample, and do not let the hunt for proof delay antifungal treatment when suspicion is high and the disease is racing.
The laboratory tests that build the diagnosis
The diagnosis is an assembly job. No single test rules in or rules out the disease. The tests below each cover part of the problem.
Direct smear and culture are the old base. A potassium hydroxide wet mount, also called a KOH wet mount, reads at 60 to 70 percent sensitivity and about 70 percent specificity. Immunofluorescence staining of deep samples reaches about 80 percent sensitivity and 97 percent specificity, so it is worth doing on every sample. Culture identifies the species and enables susceptibility testing, but it is slow, often negative, and cannot by itself tell infection from colonisation. A single negative smear does not exclude infection; sample more than once and from more than one site.
PCR depends on the sample. From a sterile site, tissue or sterile fluid, a positive PCR is strong evidence. From a non sterile site such as sputum, it is easily distorted by colonisation. For suspected invasive pulmonary aspergillosis, bronchoalveolar lavage PCR reads at 77 to 91 percent sensitivity and 80 to 95 percent specificity. Two consecutive positive blood Aspergillus PCR results push specificity to about 95 percent. For pulmonary mucormycosis, bronchoalveolar lavage PCR reaches 98 percent sensitivity and 96 percent specificity, and fresh tissue PCR beats formalin fixed tissue. For Fusarium, sample the infected site, joint fluid or sinus aspirate.
The beta D glucan test, the G test, is fast and non invasive and helps with Aspergillus, Candida and Pneumocystis, but it does not work for Cryptococcus or Mucorales. In a meta analysis of 16 studies it read 76.8 percent sensitivity and 85.3 percent specificity. In a large Cochrane review of 49 studies the between study heterogeneity was too high to pool, and in critically ill patients the sensitivity fell as low as 27 percent. A positive G test appears nearly 6 days before a positive blood culture. It cannot name the genus and should not stand alone.
The galactomannan test, the GM test, is the main test for aspergillosis and is not produced by Mucorales, so it cannot diagnose mucormycosis. Serum GM in non neutropenic critically ill patients is weak, about 30 percent sensitive. Bronchoalveolar lavage GM is far better at 75.68 percent sensitivity against 37.84 percent for serum. A meta analysis put BALF GM for invasive pulmonary aspergillosis at 87 percent sensitivity and 89 percent specificity. In viral pneumonia associated aspergillosis, BALF GM reads about 85 percent sensitive and over 80 percent specific, rising to 94 percent specific if no antifungal treatment has started. Cerebrospinal fluid GM for suspected central nervous system disease reads 57.2 to 78.7 percent sensitive and 82.8 to 98.3 percent specific.
| Test | Sample | Threshold | Use |
| BALF galactomannan (BALF GM) | Bronchoalveolar lavage | Optical density index 1.0 or above | Preferred test in critically ill patients |
| Serum GM | Serum | Optical density index 0.5 or above on a single test | A negative result does not exclude aspergillosis in non neutropenic patients |
Next generation sequencing is sensitive but not yet standardised. It cannot reliably separate pathogen from colonist from contaminant, thresholds vary, and contamination is a real risk. It is not a routine test and does not replace microscopy, serology or pathology. It earns a place when standard tests are negative or inconclusive and suspicion stays high, using sterile site samples and paired with conventional testing.
The diagnostic categories
The diagnosis is graded as possible, probable and proven.
| Evidence | Possible | Probable | Proven |
| Risk factor | Present | Present | Present or absent |
| Clinical feature | Present | Present | Present |
| Imaging | Present for lung, sinus, brain | Present for lung, sinus, brain | Present for lung, sinus, brain |
| Microbiology | Absent | Present | Present, must come from a sterile sample |
| Histopathology | Absent | Absent | Present |
Possible needs at least one risk factor and one clinical feature with no microbiology. Probable adds one microbiological item. Proven requires histology showing invasion, and any microbiology must come from a sterile source.
Treating invasive aspergillosis
For invasive pulmonary aspergillosis the first line is isavuconazole or voriconazole given intravenously, then stepped down to oral once the patient is improving and the gut works. Liposomal amphotericin B is the alternative when triazoles are not tolerated or resistance is suspected, and amphotericin B deoxycholate is the fallback when resources are thin. Posaconazole and echinocandins are supporting options; echinocandins should not be used alone.
The evidence behind the first line is consistent. A large randomised trial showed voriconazole gave 70.8 percent 12 week survival against 57.9 percent for amphotericin B deoxycholate, with less toxicity. The SECURE trial showed isavuconazole non inferior to voriconazole for first line invasive aspergillosis with better tolerance and fewer drug interactions. Across several meta analyses isavuconazole ranks best for clinical response and for reduced mortality. It also carries the lowest rate of serious adverse events.
| Drug | Role | Loading dose | Maintenance dose |
| Isavuconazole | First line | 200 mg every 8 hours for 2 days | 200 mg once daily |
| Voriconazole | First line | 6 mg per kg every 12 hours on day 1 | 4 mg per kg every 12 hours |
| Liposomal amphotericin B | Alternative | 1 mg per kg per day, raise as needed | Up to 3 mg per kg per day |
| Posaconazole | Alternative | 300 mg every 12 hours on day 1 | 300 mg once daily |
| Caspofungin | Alternative, never alone | 70 mg on day 1 | 50 mg once daily |
Treatment should run at least 6 to 12 weeks, until every sign of infection is gone. The length depends on how deep the immune suppression goes and how fast it lifts.
Central nervous system aspergillosis is the most lethal form. Voriconazole is preferred because its small molecular weight, 349 Da, and moderate lipophilicity let it cross the blood brain barrier and reach the brain and cerebrospinal fluid at effective levels. Isavuconazole also penetrates brain tissue well and is a valid alternative, though its cerebrospinal fluid levels are limited. Liposomal amphotericin B reaches brain tissue but little free drug in adult cerebrospinal fluid, so it sits behind the triazoles. For sinus and other extrapulmonary disease voriconazole remains first choice, with isavuconazole or liposomal amphotericin B for those who cannot take it.
Combination therapy is not routine for first line aspergillosis. The largest trial, voriconazole with anidulafungin against voriconazole alone, showed 19.3 percent versus 27.5 percent six week mortality, a difference that did not reach significance. Meta analyses agree there is no clear survival benefit from starting with two drugs. Combination earns its place in salvage: first line failure, multi site disease, suspected azole resistance or rapid progression. The common pair is a triazole, isavuconazole or voriconazole, with an echinocandin such as caspofungin. Salvage combination response reaches 53 percent with 69 percent 12 week survival, better than single drug salvage.
Treating mucormycosis
Mucormycosis treatment starts with liposomal amphotericin B at 5 to 10 mg per kg per day, with no slow ramp. For solid organ transplant or central nervous system involvement use the upper end, 10 mg per kg per day. Isavuconazole is an alternative and, in the VITAL trial, reached 33 percent 42 day mortality against 39 percent for liposomal amphotericin B, a non inferior signal, with no nephrotoxic cyclodextrin excipient so no dose cut is needed in kidney failure. Posaconazole by mouth suits renal impairment.
Surgery is part of the cure, not an extra. Mucorales invades vessels, causes thrombosis and infarction, and drugs cannot reach the dead tissue. For a localised, removable lesion in a patient who can tolerate it, operate early alongside antifungals. For the sickest, surgery plus antifungals beats either alone: survival was 70 percent with both, 61 percent with surgery alone and 57 percent with antifungals alone. For pulmonary mucormycosis, discharge survival was 73.2 percent with surgery plus antifungals against 35.6 percent with antifungals alone. Course length is long, at least 3 to 6 months for COVID 19 associated pulmonary disease.
Treating fusariosis and the rare moulds
Sustained immune suppression is the strongest driver of Fusarium death, so lifting it is the main lever on outcome. Fusarium is widely resistant, which narrows the options. First line is voriconazole or liposomal amphotericin B, and many teams start with voriconazole plus liposomal amphotericin B. Posaconazole is the salvage choice after first line fails. Courses run until infection is controlled and immunity is rebuilt, often months. For Scedosporium and Lomentospora, voriconazole is the preferred backbone, often with terbinafine, and surgery where possible.
Special patients and dose adjustment
Critically ill bodies do not handle drugs the way stable outpatients do. Obesity, liver injury, kidney injury, renal replacement and ECMO all move exposure.
Obesity changes pharmacokinetics through body composition and organ function. Isavuconazole needs no change and no routine drug monitoring. Voriconazole intravenous dose is best figured by adjusted body weight using a 0.4 correction factor; oral voriconazole uses standard doses. Posaconazole uses standard doses, with high dose above 120 kg. Liposomal amphotericin B is dosed by total body weight. Caspofungin needs more above 80 kg.
Liver injury calls for cutbacks by Child Pugh class. Isavuconazole needs no change until severe, where maintenance is halved. Voriconazole holds until moderate, then halves maintenance and needs monitoring. Caspofungin and micafungin drop to 35 mg per day in moderate and severe injury.
Kidney injury means watching both drug nephrotoxicity and the solvent. Isavuconazole needs no adjustment even in end stage renal disease because its intravenous form lacks the cyclodextrin solvent. Voriconazole and posaconazole intravenous forms carry cyclodextrin that accumulates, so oral is preferred and venous use needs a benefit risk call. Liposomal amphotericin B needs no dose change, but dextrose deoxycholate amphotericin B is best avoided in liver and kidney failure.
During continuous renal replacement, most antifungals need no dose change because they are cleared by the liver or bound tightly to protein. Voriconazole is the exception: it shows non linear metabolism in the critically ill and strongly warrants drug monitoring. The rule of thumb is that if clearance by the filter tops a quarter of total clearance, the dose must be reconsidered, and loading doses tied to distribution volume are kept.
During ECMO, triazoles, especially voriconazole and posaconazole, tend to fall below target concentration, while isavuconazole stays steadier. Monitor voriconazole and posaconazole, consider monitoring isavuconazole, and re assess the dose after ECMO comes off.
Antifungal drug interactions and therapeutic drug monitoring
Triazoles block CYP450 enzymes and P glycoprotein. Voriconazole strongly inhibits CYP2C9, CYP2C19 and CYP3A4; posaconazole strongly inhibits CYP3A4. That shifts the levels of calcineurin inhibitors, mTOR inhibitors, some antibacterials, anticoagulants, proton pump inhibitors and sedatives. Tacrolimus and ciclosporin doses must drop sharply; rivaroxaban is forbidden with posaconazole; rifabutin and rifampicin are forbidden with both. Amphotericin B adds synergistic kidney toxicity with aminoglycosides, vancomycin and polymyxins. The practical move is to pick the lower interaction drug, isavuconazole or an amphotericin B formulation, when feasible, watch levels and toxicity, and use the pharmacist on the team.
Antifungal therapeutic drug monitoring matters most for drugs with a narrow window and unpredictable kinetics. Voriconazole target trough is about 2 to 6 mg per L, with levels above 5.5 mg per L tied to liver and nerve toxicity; many Asian patients do better under 4.5 mg per L. Posaconazole target trough is above 0.5 mg per L for prophylaxis and above 1 mg per L for treatment, with toxicity above about 3.4 mg per L. Isavuconazole target trough is above 1 mg per L, with gut side effects above 5.1 mg per L. Caspofungin monitoring is reserved for the very sick. Model informed precision dosing builds the start dose from prior data and can simulate changes without waiting for steady state, which suits patients whose physiology swings day to day.
Reading the response
Response to invasive pulmonary aspergillosis treatment is graded as success or failure. Success splits into complete response, infection cleared, symptoms gone, imaging resolved or left with only scar, and partial response, patient alive, symptoms improved, lesion shrunk by at least 25 percent. Stable response means minimal change. Progression means worse clinically or on imaging, or a new site. The grade drives how long to treat and whether to switch or combine.
Limitations
This is a clinical overview built from published evidence and consensus practice, not a substitute for local guidance or a specialist consult. The numbers here are ranges from heterogeneous studies, and between study definitions, assays and patient mixes vary, so figures should guide suspicion rather than fix a diagnosis. Resistance patterns differ by region and change over time, so local epidemiology and antifungal stewardship must shape every regimen. Doses given are typical adult ranges and must be individualised by weight, organ function, drug levels and the pathogen.
Bottom line
Invasive filamentous fungal disease in critically ill patients is common, fast and deadly, and it no longer waits for neutropenia. Suspect it in the ICU patient with the right host and the wrong response to routine care, prove it with bronchoalveolar lavage galactomannan and PCR plus targeted tissue, and treat it early with isavuconazole or voriconazole for aspergillosis and liposomal amphotericin B with early surgery for mucormycosis. Adjust the dose to the body and the machine, watch the levels, and manage the interactions.
FAQ
What is invasive filamentous fungal disease?
Invasive filamentous fungal disease is a mould infection that spreads into tissue and blood vessels. In critically ill patients it is driven by Aspergillus, the Mucorales order and Fusarium. It starts fast, causes necrosis and vessel invasion, and carries high mortality. The lung is the most common site, though the sinuses, brain, gut and skin are also involved depending on the pathogen and the host.
Which fungi cause filamentous fungal disease in ICU patients?
Aspergillus is the most common cause, with Aspergillus fumigatus making up 50 to 90 percent of cases. The Mucorales order causes rhino orbital cerebral, pulmonary and cutaneous disease and is naturally resistant to voriconazole. Fusarium often produces fungaemia in 40 to 60 percent of patients and spreads widely. Species identification matters because it decides which antifungal to avoid before culture results return.
What are the main risk factors in non neutropenic ICU patients?
Risk comes from the host, the treatment and the environment. Host factors include severe COPD, severe influenza, severe COVID 19, uncontrolled HIV, cirrhosis and poorly controlled diabetes. Treatments include high dose glucocorticoids and calcineurin or mTOR inhibitors. Environment adds mould spore exposure from construction or contaminated wounds. Diabetes with ketoacidosis is the hallmark doorway to Mucorales.
How is invasive aspergillosis diagnosed in the ICU?
Diagnosis combines host risk, compatible imaging and microbiology. Bronchoalveolar lavage galactomannan reads about 87 percent sensitivity and 89 percent specificity, and BALF PCR adds 77 to 91 percent sensitivity. A positive serum galactomannan or beta D glucan supports the picture but is weaker in non neutropenic patients. Tissue showing invaded hyphae proves it. No single negative test rules the disease out.
What is the role of the galactomannan test?
Galactomannan is a cell wall marker released by Aspergillus and is the key nonculture test for aspergillosis. A bronchoalveolar lavage value at or above an OD index of 1.0 is the preferred threshold, with about 87 percent sensitivity. Serum galactomannan is far less sensitive at around 30 percent in non neutropenic patients. Mucorales do not release galactomannan, so the test cannot diagnose mucormycosis.
Is the beta D glucan test useful for mucormycosis?
No. The beta D glucan test, also called the G test, helps detect Aspergillus, Candida and Pneumocystis but it is not produced by Mucorales, so a negative result cannot exclude mucormycosis. In critically ill patients its sensitivity can fall as low as 27 percent, and it cannot name the genus. It is a supporting test at best and should never stand alone as proof or as a reason to withhold treatment.
What is first line treatment for invasive pulmonary aspergillosis?
First line treatment is isavuconazole or voriconazole given intravenously, then stepped down to oral as the patient improves. Liposomal amphotericin B is the alternative for intolerance or suspected resistance, and posaconazole is a supporting option. Caspofungin should not be used alone. Treatment runs at least 6 to 12 weeks.
How is central nervous system aspergillosis treated?
Voriconazole is preferred for central nervous system aspergillosis because it crosses the blood brain barrier. Isavuconazole is a valid alternative with good brain tissue penetration but limited cerebrospinal fluid levels. Liposomal amphotericin B reaches brain tissue but little free drug in cerebrospinal fluid, so it sits behind the triazoles. Surgery helps for abscess or blockage.
What is the treatment for mucormycosis?
Mucormycosis treatment starts with liposomal amphotericin B at 5 to 10 mg per kg per day, using the upper dose for transplant or brain involvement. Isavuconazole is an alternative and needs no dose cut in kidney failure. Posaconazole by mouth suits renal impairment. Courses run long, at least 3 to 6 months for COVID 19 associated pulmonary aspergillosis (CAPA).
When is surgery needed in mucormycosis?
Surgery is needed early when the lesion is localised and removable, because Mucorales invades vessels and causes infarction. Operating alongside antifungals beats either alone: survival was 70 percent with both, 61 percent with surgery alone and 57 percent with antifungals alone. For pulmonary disease, adding surgery raised discharge survival to 73.2 percent against 35.6 percent with drugs alone.
Why is therapeutic drug monitoring important for antifungals?
Antifungal therapeutic drug monitoring matters because antifungal exposure varies widely in the critically ill and the window is narrow for some drugs. Voriconazole target trough is about 2 to 6 mg per L, with toxicity above 5.5 mg per L. Posaconazole and isavuconazole also have target ranges. Monitoring lets the team individualise the dose by weight, organ function and drug interactions.
How are antifungal doses adjusted during CRRT or ECMO?
During continuous renal replacement most antifungals need no change because they are liver cleared or protein bound, but voriconazole warrants monitoring due to non linear metabolism. During ECMO, voriconazole and posaconazole tend to fall below target while isavuconazole stays steadier. Re assess the dose after the machine stops.
